In a watershed achievement, the Human Genome Project (HGP) recently sequenced the entire human genome, providing a wealth of information about potential genes and regulatory sequences. Despite this success, exactly how genomic sequence specifies the behavior and development of complex organisms remains largely unknown. Gene expression within cells is tightly regulated, with many genes expressed only under certain environmental conditions or at stereotyped time points during development. The next great challenge lies in developing a mechanistic understanding of how regulatory sequences dictate gene expression, with the ultimate goal of being able to quantitatively predict expression levels from sequence. Solving this challenge would have far-reaching impacts in biology, elucidating how changes in regulatory sequence can lead to transcriptional dysfunction and disease and improving rational design of transgenes for gene therapy. Regulation of gene expression is accomplished primarily via binding of transcription factors at specific genomic loci. Once bound, transcription factors can either recruit or block the general transcription machinery, thereby activating or repressing transcription. Most leading models of transcriptional regulation are built upon thermodynamic principles, and require information about transcription factor concentrations in vivo and their affinities for different DNA sequences. Despite this central role for binding affinities, experiments to date have been forced to infer affinities from genome-wide occupancy and expression measurements due to a lack of biophysical data. Using a recently developed microfluidic system that permits the high-throughput measurement of interaction affinities, this proposal seeks to systematically investigate the thermodynamics of transcriptional regulation at multiple scales, from individual interactions between transcription factors and target sequences to the nucleation of assemblies of DNA binding proteins at regulatory loci. Experiments will focus on, in turn: (1) how particular contacts between protein residues and DNA bases determine interaction affinities; (2) how cell-specific signals modify these interactions to dictate tissue-specific expression patterns; (3) how evolutionary changes in both regulatory DNA sequences and transcription factors rewire transcriptional networks during evolution to drive phenotypic change; and (4) how cooperativity and competition between transcription factors affect binding patterns to influence gene expression. Data from these experiments will provide crucial information required to construct ground-up, quantitative models of transcriptional regulation and increase our ability to predict gene expression from regulatory sequence. The funding provided by this K99 award would provide crucial resources for the PI, Polly Fordyce, to receive 2 years of additional formal training in the biological sciences and ensure a successful transition to an independent career.